DE69508887T2 - Control system for the inclination of a railroad car body - Google Patents

Description

The present invention relates to a system for controlling the rotation of a railway carriage body.

In particular, the present invention relates to a control system for rotating the car body about a longitudinal axis of the car to reduce the acceleration to which the passengers are subjected.

Railway carriages with variable trim are known which, when cornering, allow the carriage body to rotate around a longitudinal axis of the carriage in order to compensate for the centrifugal lateral acceleration to which the passengers are subject by the gravitational acceleration.

When cornering, passengers in a railway carriage with variable trim are therefore subject to a lateral force equal to the difference between the transverse centrifugal force and the transverse weight force.

Known control systems contain (e.g. hydraulic) actuators for controlling the inclination angle of the car body based on a drive signal.

More specifically, control systems are known that provide control circuits for generating the drive signal according to predetermined laws and on the basis of a number of input parameters measured in the car.

Known control circuits include electronic circuits for generating a reference signal representing a reference car body tilt angle pattern based on a number of input parameters.

The reference signal is compared with a signal indicating the actual tilt angle of the car body, with the resulting error signal being used in a closed loop to calculate the drive signal.

During the transient conditions at the beginning and end of the curve, known control systems cannot achieve accurate convergence of the reference and actual pitch angle signals, so that the actual pitch angle of the car often deviates significantly from the reference angle, resulting in a lateral force sufficient to impair passenger comfort.

It is an object of the present invention to provide a control system designed to eliminate the disadvantages typically associated with known systems.

According to the present invention, there is provided a control system for the rotation of a railway car body according to claim 1.

The present invention is described with reference to the accompanying drawings, in which:

Figure 1 shows a schematic cross-section of a railway carriage having a control system according to the present invention;

Fig. 2 shows a schematic representation of an electronic circuit implementing the control system according to the present invention;

Figures 3 and 4 show graphs of a number of quantities of the system according to the present invention;

Fig. 5 shows a schematic representation of an electronic circuit with which a known control system is implemented.

The reference numeral 1 in Fig. 1 indicates a control system applied to a railway carriage 3 with variable trim, which essentially comprises two or more bogies 5, each of which is elastically connected via a suspension 7 to axles 8 fixed to wheels 9 running along rails 10, and a carriage body 11 which is inclinable about a longitudinal axis G of the carriage 3 by means of a hydraulic device 12.

The device 12 essentially contains a right and a left hydraulic actuator 13r, 13l, which are inserted between the car body 11 and the chassis 5, as well as a servo valve 15 for supplying the actuators 13r, 13l.

When cornering, the passengers (not shown) are subject to an uncompensated lateral acceleration Anc, as shown schematically in Fig. 1, which is roughly given by the following equation:

Anc = V² / R - gφ (1)

where V is the speed of travel of the carriage 3, R is the curve radius, g is the gravitational acceleration and φ is the angle between the support surface P of the rail 10 and the horizontal plane H.

The uncompensated acceleration Anc can be reduced by rotating the car body 11 by an angle θ around the centre of gravity axis G as shown in Fig. 1, so that Anc is given by:

Anc = V² / R - g(φ + θ) (2)

which is obviously smaller than the value in (1).

The system 1 includes an electronic control unit 14, which is supplied with a number of parameters measured in the car 3 and which generates a drive signal for the servo valve 15 (shown schematically). The servo valve 15 has an inlet 15a supplied with pressurized fluid 16 and supplies with its output the hydraulic actuators 13r, 13l to adjust the tilt angle 9 of the car body 11.

More specifically, the servo valve 15 has a first outlet 15r, which communicates with the actuator 13r via a circuit 17r; a second outlet 15l, which communicates with the actuator 13l via a line 17l, and a return outlet 18u.

The servo valve 15 includes a hollow outer casing (not shown) housing a central sliding valve (not shown) that is axially movable in the straight direction X by a pressure difference dependent on a drive signal supplied to an electrical driver 21.

More specifically, the servo valve 15 is of the open center type, i.e., when the slide valve is located centrally (X = 0) and the driver 21 is not supplied, the outlets 15r, 15l communicate with the return outlet 18u.

The sliding valve is also movable by the driver 21 between a left limit position in which the inlet 15a communicates with the outlet 15l and the outlet 15r communicates with the return outlet 18u, and a right limit position in which the inlet 15a communicates with the outlet 15r and the outlet 15l communicates with the return outlet 18u, movable.

When the servo valve 15 moves to the left limit position, the pressurized fluid is supplied to the actuator 13l while the actuator 13r is exhausted to rotate the car body 11 clockwise; when the servo valve 15 moves to the right limit position, the pressurized fluid is supplied to the actuator 13r while the actuator 13l is exhausted to rotate the car body 11 counterclockwise.

Reference numeral 22 in Fig. 2 designates a control circuit of the control unit 14 in accordance with the teachings of the present invention.

The control circuit 22 comprises a first node 23, which has an addition input 23a and a subtraction input 23b. The addition input 23a is supplied with a reference signal θc generated by a circuit 26 (e.g. an electronic card) supplied with a number of signals P1, P2, ..., Pn relating to parameters measured in the carriage 3 (e.g. speed and acceleration of the carriage).

The signal θc represents the ideal pattern (Ref) of the angle θ as a function of time t, as shown by the Ref curve in Fig. 3, which has a first section A (between time t = 0 and time t = T1) in which the angle θ is constantly increasing and a second section B in which the angle θ assumes a constant value θlim (for t > T1).

The subtraction input 23b is supplied with a feedback signal θz, which is generated by a sensor 28, which measures the instantaneous angle θ of the car body 11.

The node 23 has an output 23u which communicates with the input 31a of a circuit 31 which has a constant proportional gain Gp and generates a signal equal to the product of the input signal and the proportional gain Gp.

An error signal θe = θc - θz is present at the output 23u, which represents the error between the angle θ (θc) requested by the control system and the actual angle (θz) of the car body 11 of the car 3.

The circuit 31 has an output 31u which communicates with a first addition input 34a of a node 34, which has an output 34u at which a signal I is present, which, when amplified, forms the drive signal of the driver 21.

According to the present invention, the control circuit 22 includes a differentiating circuit 37 having an input 37a communicating with the output 23u of the node 23 and an output 37u communicating with a second addition input 34b of the node 34. The circuit 37 generates a signal equal to the differential of the input signal multiplied by a derived gain Gd.

The differential gain Gd of the circuit 37 depends, as shown in Fig. 4, on the time derivative of the absolute value (d θe /dt) of the error signal θe and has three sections:

- a first section (K) between a first value (0) and a second value (e.g. 0.2 degrees/second) of the derivative of the absolute value of the error signal (d θe /dt), wherein the differential gain Gd is zero;

- a second section (L) between a second value (e.g. 0.2 degrees/second) and a third value (e.g. 2.2 degrees/second) of the derivative of the absolute value of the error signal (d θe /dt), the differential gain Gd constantly increasing with a constant slope; and

- a third section (M) in which the derivative of the absolute value of the error signal (d θe /dt) exceeds the third value (e.g. 2.2 degrees/second), the differential gain Gd assuming a constant value (0.5 mA/(º/s)).

The operation of the circuit 22 will now be described with reference to Figs. 2 and 5, beginning for the sake of clarity with a description of the operation of known control systems.

A first known type of (proportional) control system includes a node 23 (Fig. 5) supplied with a reference signal θc indicating the ideal pattern of angle θ and a feedback signal θz indicating the actual angle θ of the car body; node 23 supplies an error signal θe equal to the difference between the reference signal and the feedback signal to a proportional circuit 31 having a constant proportional gain Gb; further, circuit 31 generates a drive signal which, when converted and amplified, controls the solenoid valve supplying the hydraulic actuators to adjust the inclination angle θ of the car body.

The reference signal can be as shown by the Ref curve in Fig. 3 and contains a first section A in which the angle θ increases continuously, and a second section B in which the angle θ assumes a constant value θlim .

During the transients at the beginning and end of the curve, proportional control systems cannot produce an effectively "tracked" reference signal Ref, so that for physical reasons not discussed here, the actual angle θ of the car body deviates significantly from the reference angle, as shown by the dashed-dotted curve Mis in Fig. 3, which indicates the actual angle pattern as a function of time. As can be seen, at the beginning of the time scale (section W), i.e. in the initial control stage, the Mis curve deviates significantly from the Ref curve and approaches it when the reference signal Ref assumes a constant value (section V). As a result, the uncompensated acceleration in the initial control stage is far from negligible, to the extent that it affects passenger comfort.

To eliminate the above disadvantage, proportional/derivative control systems have been used which have a differentiating circuit D (shown by the dashed line in Fig. 5) in parallel with the proportional circuit 31 and produce a signal proportional to the differential of the error signal. Since this is positive in section W and negative in section V, the differentiating circuit reduces the error in section W, but increases it in section V by adding to the Mis curve a positive term (error differential greater than zero) in section W and a negative term (error differential less than zero) in section V. The angle θ signal obtained using the proportional/derivative systems is shown by the dashed-dot curve Ref2 which has a significant error increase in section V.

The control circuit according to the present invention effectively eliminates the disadvantages of both proportional systems and proportional/derivative systems.

As shown in Fig. 3 and in the circuit of Fig. 2:

- in the section W (where the error signal differential is positive), the gain Gd of the circuit 37 increases along with the error differential, as shown in Fig. 4, so that the positive signal generated by the circuit 31 is added to the signal (Mis) generated by the circuit 31 (at node 34), thereby reducing the total error;

- in section V (where the error signal differential is negative or slightly positive) the gain Gd of the circuit 37 is zero (or very small) as shown in Fig. 4, so that to the signal (Mis) generated by the circuit 31 (at node 34) is added a small signal (or zero signal) with very little effect on the Mis signal, so that the error remains small.

The control device 22 therefore provides an effective "tracking" of the reference signal by maintaining a very small error in both sections W and V between the reference signal and the feedback signal, so that the angle actually assumed by the carriage approaches very closely to the reference angle.

The control circuit 22 also includes an integration circuit 40 having an input 40a connected to the node 23 via a threshold circuit 42 and an output 40u connected to an addition input 34c of the node 34 via a limiting circuit 44.

The circuit 40 forms the integral of the input signal and multiplies it by an integration gain Gi; the Circuit 42 is a threshold comparison circuit which blocks any input signal whose absolute value is below a threshold value ε.

The operation of the circuit 40 will now be described with reference to the physical operation of the servo valve 15.

The sliding valve of an ideal servo valve is theoretically set to the center position (X = 0; hydraulic-zero) when no drive current is sent to the driver 21 (electrical-zero), so that fluid is supplied to the actuators 13r, 13l at the same pressure.

However, in real practice, for various physical reasons (e.g. temperature change of the servo valve components, acceleration of the servo valve, machining tolerances), hydraulic zero does not correspond to electrical zero. That is, in a real servo valve (i.e. a servo valve of the type fitted in railway carriages), hydraulic zero is achieved with a non-zero drive current (bias current) and does not correspond to zero drive current under any circumstances.

Therefore, if the control circuit 22 were not provided with the integration circuit 40, a zero drive current would be generated when the error is eliminated (equilibrium position of the system), so that the servo valve 15 would not be placed in the hydraulic zero position for the above reasons, the pressure of the actuators 13r, 13l would not be the same and a position error other than zero would remain.

However, the circuit 40 eliminates the above disadvantage by continuously forming the integral of the error signal θe and by generating a signal that continuously increases as long as the error signal θe is present. If the equilibrium position is reached, the error is zero and the output signals of the circuits 31, 37 are zero, but the node 34 is supplied with the output signal of the integrator 40.

Thus, a drive current is generated which, by appropriately dimensioning the gain Gi, forms the bias current for the correct positioning of the sliding valve and the achievement of hydraulic-zero in the equilibrium position of the system.

Of course, changes may be made to the control system as described and shown without departing from the scope of the present invention.

Claims (6)

1. System for controlling the rotation of a railway car body in order to reduce the uncompensated acceleration to which the passengers are subjected; the railway car comprising a car body (11) in which the passengers are accommodated and a rotating device (12, 13r, 13l, 15, 14) for rotating the car body (11) at least about a longitudinal axis (G) of the car (3); wherein the rotating device (12, 13r, 13l, 15, 14) rotates the car body (11) in response to a drive signal (I) by a given angle (θ); wherein the control system (1) contains an electronic computing device (14, 22) for generating the driver signal (I); wherein the electronic computing device (22) contains: - reference means (26) for generating a reference signal (θc) indicative of the required pattern of the angle; - a first node device (23) which is supplied at its input (23a) with the reference signal (θc) and with a feedback signal (θz) which indicates the actual value of the angle; wherein the first node device (23) generates at its output an error signal (θe) which is equal to the difference between the reference signal (θc) and the feedback signal (θz); - a proportional device (31) which is supplied with the error signal (θe) and has at its output (31u) generates a signal equal to the product of the input signal (θe) and a proportional gain (Gp); - a differentiating device (37) having an input (37a) supplied with the error signal (θe) and producing at its output (37u) a signal equal to the differential of the input signal (θe) multiplied by a differential gain (Gd); - a second node device (34) which is supplied with the output signal (31u) of the proportional device (31) and with the output signal (37u) of the differentiating device (37); wherein the second node device (34) generates the driver signal at its output (34u); characterized in that the differentiation device (37) has a differential gain (Gd) which depends on the time derivative of the absolute value (d θe /dt) of the error signal (θe). 2. System according to claim 1, characterized in that the differential gain (Gd) decreases together with a reduction of the time derivative of the absolute value of the error signal (θe). 3. System according to claim 1 or 2, characterized in that the differential gain (Gd) has a pattern that contains at least three sections: - a first section (K) between a first value (0) and a second value (0.2 degrees/second) of the derivative of the absolute value of the error signal (d θe /dt), in which the differential gain (Gd) is substantially zero; - a second section (L) between the second value (0.2 degrees/second) and a third value (2.2 degrees/second) of the derivative of the absolute value of the error signal (d θe /dt), in which the differential gain (Gd) continuously increases with a constant slope; and - a third section (M) in which the derivative of the absolute value of the error signal (d θe /dt) exceeds the third value (2.2 degrees/second) in which the differential gain (Gd) assumes a constant value (0.5 mA/ (º/s)). 4. System according to any one of the preceding claims, characterized in that the reference device (26) generates a reference signal (θc) which indicates the ideal pattern of the given angle (θ) as a function of time (t); the pattern comprising a first section (A) in which the angle (θ) increases continuously and a second section (B) which is adjacent to the first section (A) and in which the angle (θ) assumes a constant value (θlim). 5. System according to any one of the preceding claims, characterized in that it includes an integration device (40) comprising an input (40a) communicating with the output (23a) of the first node device (23) and an output (40u) communicating with an input (34c) of the second node device; the integration device (40) generates at its output (40u) a signal which is equal to the integral of the input signal (θe) multiplied by an integration gain (Gi). 6. System according to any one of the preceding claims, characterized in that the rotating device (12, 13r, 13l, 15, 14) contains at least a first and a second actuator (13r, 13l) for rotating the car body (11) in opposite directions; and a servo valve (15) for supplying the actuators (13r, 13l); wherein the servo valve (15) has at least one inlet (15a) supplied with pressurized fluid (16); and a first and a second outlet (15r, 15l), which communicate with the first and second actuators (13r, 13l), respectively; characterized in that the servo valve is of the open center type.